Hacker News Reader

Comment by doki_pen | Hacker News Reader

It seems that the rules are pretty well defined. Why can't computers auto-kern?

The other comments here are true, but it's worth pointing out explicitly: High-quality fonts have already had this done to them, and contain the resulting information explicitly in the font file. Most software packages (Latex since forever, recent versions of MS Word) understand how to read this information and apply it automatically to text. There are relatively few cases were you actually need to touch this in practice, assuming your fonts and software are doing their job.

The article itself is describing what you would do to fix kerning in a low-quality font. But really, why would you want to waste your time with that? Even properly kerned, a low-quality font will have other flaws.

There are still certain typographical tasks like avoiding rivers (https://en.wikipedia.org/wiki/River_%28typography%29) which are not handled well by existing typographical engines. But most everything else can be done well with Latex and a good font.

So the real lessons here are:

1. Pay for high-quality fonts. This is just one more reason to do so.

2. If you have to use MS Word, use a recent version or run it through Adobe InDesign afterwards. Or just do it correctly from the beginning with Latex.

3. For very large text that will be used frequently (such as in logos), pay for a designer.

This is called "Optical" kerning in Adobe products. In general it's a good time saving device if for some reason you're using a typeface with no kerning pairs defined, but the result often ends up looking robotic and either too tightly or too loosely spaced (as I believe it ends up kerning every pair rather than only those that need kerning). If you have a typeface made by a reputable foundry there's usually no need to turn this feature on.

On the type design end of kerning, automated kerning does exist, though it's always tuned by hand afterwards. The most complicated part of kerning is that there are variables that need to be tuned based on the purpose of the typeface. If the typeface is intended for screen use or print use, or if it's made for large headlines or tiny captions, the amount and method of kerning changes, and so would the algorithm needed to generate it.

I can't imagine how kerning could look 'robotic'. Can you give an example?

Not an easy phenomenon to describe or even illustrate. I believe that some legibility come through slightly uneven textures that are created by varying shapes that words create and the "color" (density of light and dark) that is created by different sequences of letters. Different words become identifiable by their shape and the visual space they occupy. Kerning somewhat counteracts this by attempting to make sequences appear more regular, so "robotic" kerning, might be described as an extemely even appearance to the detriment of legibility.

Granted were talking about a very micro adjustment. In the end he word will still be word and will still be read.

That sounds like a perfect match for using machine learning. The exact connection what we think is the perfect kerning isn't obvious and the model used by Adobe seems to be arbitrarily selected. Machine learning probably can approximate how a human decides which kerning is best better.

Right, though I'm skeptical that the program would also be able to correctly identify the motivation for the typeface, which drives all of the decisions about how it behaves. That component of the program would end up being far more complex than generating the actual kerning pairs themselves.

They can. Most software (MS Word) will just use kerning pairs predefined in the font, but more complex design software (Adobe InDesign and Illustrator, for example) can automatically optically kern text - it works pretty dang well.

Maybe I'm not picky enough as a designer, but I only very rarely kern something by hand, and that's usually when I'm working with some weird display font that auto kern doesn't handle well and has poorly defined kern pairs. Sometimes logos as well, but for 99% of text it's really not worth the trouble IMHO.

You're probably not gonna kern by hand anything that isn't an attention grabber in a huge font, or a logo or product label, and the like.

If you want to produce a nicely typeset document, LaTeX (or XeLaTeX) is the way to go -- it does almost everything you could want (kerning-wise) automatically, and has an active community supporting new packages and new fonts. There are also many templates[1] to help you get started.

[1] https://www.overleaf.com/latex/templates

(and if you give Overleaf a try let me know what you think - I'm one of the founders, and always like to hear feedback from new users)

No, the article is about kerning a typeface, not simply rendering existing kerns contained in font files. LaTeX does many things but auto-kerning is not one of them. It can only execute embedded kerning pairs (metric kerning) it can't infer its own (optical kerning).

Aha, thanks for that explanation, that was really useful and concise. I've learnt something new today, which is always good :)

I wondered the same thing myself. I suspect they can/do, but that their guesses aren’t great for general, more complicated fonts. That might make it inappropriate for larger print work.

Reply on news.ycombinator.com